Cooperation or Competition?

Institutional Development of Microfinance in Southern Africa

Jens Reinke*

1. Introduction

In its infancy, the microfinance movement was proud of creating savings and credit

institutions that were completely different and totally separate from those in the commercial

banking sector. Microfinance institutions were meant to be grassroots oriented, community-

based, their working practices defined by local needs, not standard banking practice

(Fuglesang and Chandler, 1986). Since then, thinking on microfinance has evolved to

emphasise financial sustainability more than before, and MFIs have come to be seen as some

form of bank themselves – unusual banks in many respects, but banks nonetheless (Otero and

Rhyne, 1994; Hulme and Mosley, 1997). Not surprisingly, the dichotomy of the early days is

retreating. Banks ‘gear down’ to provide services to small and medium business and low-

income consumers. NGOs ‘upgrade’ by adopting management practices from the corporate

sector and acquiring banking licenses. In some cases, microfinance providers get linked into

the formal financial system through cooperation with commercial banks (Seibel, 1989;

McGuire and Conroy, 1997), in some cases through donor-supported credit guarantee

schemes (Gudger, 1998).

In most respects, the microfinance sector in Southern Africa lags behind the developments in

other parts of the world. However, some of the most successful microfinance institutions in

* This article is based on the research work carried out in cooperation with the Small Enterprise Promotion Advisory Council (SEPAC). The project was financially supported by DFID.

1

Africa develop and produce their services jointly with commercial banks. Such cooperation –

known as ‘linkage’ (Soyibo, 1996) – has allowed the adoption of innovative and sometimes

unusual microfinance models, such as the minimalist model employed by the Start-Up Fund

in South Africa (Reinke, 1998a). There are, potentially, many reasons why linkages could

play an important role in developing microfinance in Southern Africa. First, microfinance

services depend on a cheap and reliable way for transferring money, such as the branch

network and the electronic transfer systems operated by commercial banks. Second, deposits

are an important aspect of microfinance services, yet few microfinance institutions qualify for

a banking license. Third, banks can help capitalise microfinance institutions through loans

and equity investment. Finally, joint training and marketing are additional possibilities.

Although the most successful examples of linkage development in Africa stem from South

Africa, financial liberalisation and institutional diversification in the financial sectors across

the region have made linkage development a potentially tenable strategy. This paper sets out

a cautious view on ‘traditional’ linkages, but finds institutional developments are turning

towards full integration of microfinance into commercial banking. In Africa, microfinance

services are mostly produced by NGOs which are neither seen nor understand themselves as

part of the financial system yet. Elsewhere, the transformation of MFIs into fully licensed

commercial banks is accepted as the normal course of graduation (Mosley, 1996). This

process has began in Southern and Eastern Africa, inter alia, with the creation of the K-REP

bank in Kenya, the Centenary Bank in Uganda and SOCREMO, a licensed non-bank

financial institution in Mozambique.

Between the given options of gearing down, grading up and establishing strategic linkages

between MFIs and banks, policy makers, donors and microfinance managers confront a wide

2

variety of institutional choices. Decision makers need to understand the purpose, role, and

limitations of linkages between microfinance institutions and banks, as well as the longer

term process of moving microfinance into the financial mainstream. This paper sets out the

experience with mainstreaming microfinance looking at the private sector developments in

Zimbabwe and the policy approaches in Tanzania.

In section 2, a conceptual overview is followed by a textbook example of ‘linkage building’

from Mozambique and an illustration of ‘intermediate’ institutions in Ghana. In section 3, the

focus is on the relationships between institutions providing financial services to micro and

small enterprises. This shows that there can be a conflict of interest as far-reaching linkage

development becomes untenable when MFIs and banks position themselves as potential

competitors. Mainstreaming microfinance implies, of course, also a role for government.

Section 4 looks at the efforts made by the Bank of Tanzania, Tanzania’s central bank, for

establishing an appropriate regulatory framework, but finds that the lack of private sector

(voluntary or profit-making) participation stymies these efforts. Section 5 concludes and

proposes an agenda for debate on mainstreaming in Southern Africa.

2. Conceptual issues: NGOs, banks, and their diminishing differences

Commercial banking was started in SADC countries under colonial rule and with the

financial needs of colonial merchants, farmers and petty capitalists in mind. Although

banking services were broadened under colonial regimes to include non-white customers –

such as Arab and Indian traders in East Africa – banks came under scrutiny for their supposed

racial bias after independence. In several countries, banks were nationalised or government

set up ‘African’ banks during indigenisation drives. However, the focus of their business

3

remained with larger companies, urban private clients, and the parastatal and government

sector, still excluding many small businesses. With privatisation and financial liberalisation

since the mid-1980s, the pressures to concentrate on profitable target groups assumed greater

importance, leading to the reiteration of claims that banks are biased towards the powerful

and wealthy. In contrast, NGOs became involved in financial dealings following the critique

of government-directed rural credit programmes. This critique not only focused on the

economic inefficiency of such lending, but also on the regressive distributional effects

(Adams et al, 1984; Braverman and Guasch, 1994). The approach of NGOs assumed a

poverty focus, aimed at rural borrowers in most cases. When NGOs dedicated themselves to

lending, or set up special poverty lending institutions, these became known as microfinance

institutions that were seen as separate from other, non-financial activities (Fuglesang and

Chandler, 1986). With an increased pressure for financial performance – ‘sustainability’, as it

became known – many microfinance NGOs moved on to leave their voluntary, welfare-

oriented touch behind (Churchill, 1997).1 Professional microfinance institutions became

increasingly serious partners for collaborators outside the voluntary sector.

Despite their move towards professionalism in finance, most microfinance institutions remain

different from banks in some important aspects. In addition to internal factors – issues such as

recruitment, pay scales and management methods – most microfinance institutions are also

less well capitalised, less well developed institutionally, and without a banking license. On

the other side, microfinance organisations are arguably less bureaucratic than banks, more

innovative (partially because they are not subject to the same scrutiny by their clients, equity

owners and regulators) and have superior knowledge of the micro-level niche market. These

1 The quest for sustainability is, of course, not motivated by donors’ avarice. In fact, continued subsidy-dependence would keep microfinance institutions permanently in competition with other ‘worthy causes’ and would limit the potential number of beneficiaries.

4

different but compatible strengths could, in principle, be the foundation for cooperation in

providing innovative, professional services to the microenterprise sector.

In South Africa, the use of commercial sector payment systems has been an important factor

in the success of innovative microfinance institutions and commercial moneylenders.

Especially the low costs of electronic payments and ATM network withdrawals, reduced the

overhead costs associated with maintaining networks of cash carrying loan officers (Reinke,

1998a). In all countries, banks maintain payments systems that may include technologically

advanced components, but also cheque clearing facilities, and branches for cash and paper-

based transactions. Efficient payment systems are important from the institutional perspective

for savings collection, loan disbursal, and repayment collection, while clients benefit not only

from convenient savings and repayment models, but potentially also from being able to use

non-cash payments for ordering stock. The economies of capital-intensive payment systems

depend on their volume, therefore most NGOs are unable to set up competing systems.

Banks are licensed as deposit-taking institutions while NGOs, in most cases, are not.

However, savings are often seen as a crucial component in microfinance packages (Owens

and Wisniwski, 1999). Finally, banks can be equity partners providing both capital and

expertise to microfinance organisations that have reached a higher level of

commercialisation. A range of possibilities is illustrated by the following case study.

GAPI/BPD: piggy-backing small enterprise finance on commercial banking

In microfinance, as in other sectors, Mozambique is often described as promising, but very

underdeveloped. A recent study found 25 microfinance institutions active in the country, but

conceded that ‘the industry is still in its infancy’ (Chidzero et al, 1998, p. 13). In addition to

5

the NGO-dominated microfinance sector, two small donor-financed institutions provide loans

to small and medium (as opposed to micro) enterprises. One of those two – GAPI – operates

in cooperation with a commercial bank. The chosen form of cooperation is a textbook

example of ‘classical’ linkage development. When established as a limited company in 1990,

GAPI was majority owned (70 percent) by Banco Popular de Desenvolvimento, a

government-owned commercial bank.2 The remainder of the capital was held by a German

donor organisation. More important than formal ownership is the cooperation in day-to-day

business. GAPI has its own offices in Maputo, but in all other regions it is represented by the

branches of the Banco Popular de Desenvolvimento. In each of Mozambique’s provinces, one

branch has one credit officer who is trained and partly paid by GAPI and answerable to both

participating organisations. As these provincial representatives conduct all interaction with

clients, assess applications and follow up loans, GAPI achieved regional outreach without the

costs of setting up its own branches. Payments are processed through the bank’s settlement

system and consolidated with GAPI.

Neither GAPI nor the Banco Popular de Desenvolvimento are entirely satisfied with the

present set-up. The privatisation of the bank has led to delays in GAPI’s transformation into a

registered financial intermediary. GAPI, although advanced by Mozambican standards, lacks

appropriate data tracking and processing systems. However, the cooperation between the two

institutions has arguably helped GAPI to expand its still very small loan portfolio, achieve a

level of regional diversification and expose its staff to banking sector working practices. Its

form of ‘linkage’ with the commercial banking sector successfully covered several areas –

equity participation, training, marketing and payment systems – that seem to be of

considerable use for other microfinance institutions. However, the benefits to the partner

2 When the Banco Popular de Desenvolvimento was privatised several years later, its stake in GAPI had to be taken over by government for legal reasons. The cooperation between the bank and GAPI continued as before.

6

bank were less clear. Following its privatisation, Banco Austral, as it is known know, reviews

its business strategy and considers terminating its relationship with GAPI, while GAPI

develops other marketing alliances and even explores possibilities for setting up a

commercial bank. This illustrates that linkages will ultimately not be sustainable if they do

not produce visible benefits to both partner organisations.

The logical end of commercialisation – graduation into the commercial banking sector –

poses questions relating the potential for longer-term relationships between banks, MFIs and

banking regulation. The appropriate legal context for licensing and supervising microfinance

banks needs to pay attention to the nature of their business. Despite its problems, the rural

banking sector in Ghana illustrates that appropriate regulatory provisions can support the

development of microfinance delivery through private-sector banks.

Microfinance banks: Ahantaman Rural Bank in Ghana

Rural banks, which account for almost all microfinance supply in Ghana, are a special feature

in this country’s financial landscape. Privately owned and for-profit, licensed and supervised

under the Banking Act, they have developmental objectives and employ the corresponding

rhetoric. There are 132 licensed rural banks in Ghana, but the Bank of Ghana classifies only

about half of them as financially sound. The greatest obstacles for rural banks are, however,

secondary reserve requirements that effectively deplete their liquid resources, leaving little

for intermediation, and the regulatory requirement that they are active only in one rural

locality. This prevents them from leveraging rural business with urban, and from realising

scale economies by geographical expansion. However, rural banks are the backbone of

microfinance delivery in Ghana. They are located between the NGO sector and fully fledged

commercial banks in terms of capitalisation, regulatory requirements and internal culture.

7

Table 1 – core data of GAPI and Ahantaman Rural Bank

GAPI Ahantaman Rural Bank Loan size Currently range from approximately

US$400 to US$300,000 Average about US$20,000

Maximum US$210

No of clients 250

2,742

Savings No savings products offered Minimum US$0.21 per day Average balance US$37

No. of savers - 6,212 Sources: private interviews and CARE Ghana. Note: figures for Ahantaman Rural Bank refer to the susu programme only. Data for GAPI stem from May 1999 while data for Ahantaman Rural Bank refers to end-1998.

Ahantaman Rural Bank, founded in 1983, set up its first microcredit programme in 1989,

based on solidarity group lending to market women and farmer groups. In 1995, it started a

microsavings mobilisation scheme called Ahantaman Susu Programme, referring to the

ambulant savings collectors known as susu in West Africa. While traditional susu collectors

are independent and work at their own risk, Ahantaman Rural Bank’s susu collectors are

wage-earning extension officers responsible for savings collection and credit extension. At

the end of 1998, the scheme had 6,212 clients (230 clients per susu officer on average), of

which over 90 percent are women. While the susu scheme also pays interest, unlike

traditional susu collectors, there are also other incentives for keeping up the savings balance.

These incentives include a lottery model where random prizes are distributed amongst those

with stable savings balances.

Clients of the susu collectors scheme have access to loans, too. Combining its various

microcredit product lines, Ahantaman Rural Bank has a loan portfolio of 2,742 loans under

500,000 cedi (US$210). It charges an interest rate of 44 percent flat. The delinquency rate

stands at 4.5 percent (overdue amounts over outstanding portfolio) and the long-run default

rate at 1.4 percent. Its performance is therefore respectable by the standards of the local

8

microfinance industry. In addition to the susu programme, Ahantaman Rural Bank provides

financial services other than microfinance. In total, it had 26,634 savers at the end of 1998;

these had deposits of over US$1.4 million. There were also 2,842 loans (including

microfinance) and 1,795 overdrafts, producing a combined portfolio of US$0.8 million.

Ahantaman Rural Bank claims to be profitable and 91 percent of its assets are commercially

funded.

Although the rural banking sector in Ghana is in need of reform and stricter prudential

requirements, it also demonstrates that intermediate institutions – no mere NGO, not quite a

fully fledged commercial bank – can play an important role in delivering microfinance

services. Regulatory requirements should enable MFIs to acquire desirable characteristics of

banks while maintaining the prerequisites for their target-group orientation. Lower

requirements on capitalisation and financial reporting as well as a more liberal regime for

unsecured lending (or the use of non-traditional forms of security) are feasible in exchange

for a narrowly defined range of products. As with other forms of banks, stability and the

security of deposits should be tantamount. Unfortunately, the regulation of rural banks in

Ghana is a legacy from pre-microfinance days and is in urgent need of modernisation.

3. Up-grading and down-gearing: commercial banks in microfinance

This section shows that the potential for cooperation between banks and NGOs dwindles and

competitive pressures set in when banks start testing the microfinance market for themselves

and NGOs develop strategies towards transforming into a formal financial institution.

Although the potential microfinance market in most developing countries is vastly larger than

the capacity of MFIs to serve it, there are scarce resources, including qualified staff, clients

9

with a track record and of course donor funding, which microfinance providers compete for.

The emergence of competition on the microfinance supply side is in principle desirable, but it

also raises new problems for microfinance managers, donors and policy makers regarding the

institution building processes of MFIs and appropriate market regulation.

Two case studies from Zimbabwe illustrate the emergence of competition. A commercial

bank has set up a new department for marketing microfinance products and becomes one of

the largest microfinance institutions in the country, helped by donor subsidies. On the other

side, the largest microfinance NGO in Zimbabwe is undergoing a process of

professionalisation in order to acquire a banking license and compete with banks head-on.3

The two institutions are not only significant for their size in the Zimbabwean microfinance

market, but they spearhead the process of moving microfinance into the mainstream in

Zimbabwe.

Commercial Bank of Zimbabwe Ltd.

The Commercial Bank of Zimbabwe is a former government-owned bank, privatised in 1997,

and one of the largest in Zimbabwe. Facing increased competition in the commercial banking

sector, management identified the hitherto neglected microfinance sector as one area where

expansion is possible. Accordingly, a microfinance programme was launched in 1996 under

the label of Community Banking. The Commercial Bank of Zimbabwe operates its

microfinance scheme under a tripartite agreement. Its partners are DFID, providing financial

assistance, and Care International for technical assistance. This scheme is a typical micro-

level lending project with low credit amounts – the average outstanding loan amount is Z$

3,170 (US$ 80) - using group solidarity structures. By April 1999, there were 1,420 active

10

loans outstanding for a loan value of Z$4.5 million. At the same time, there were also 3,400

savers in the scheme with Z$ 4.8 million in deposits. All loan officers have more than 350

clients, in one case even 437.

Table 2 - core data of donor-supported micro- and small business finance providers in Zimbabwe

Commercial Bank of Zimbabwe

Zambuko Trust Barclays Bank

Type of programme

Extensive microfinance programme within commercial bank receiving substantial donor support

NGO-type microfinance institution with Christian background, in process of graduation

Small business lending unit within commercial bank based on donor-financed credit guarantee schemes

Loan size Average US$80 Maximum US$1,000

Minimum US$50 Maximum US$2,500

Range US$200 – 150,000 Average US$1,500

No. of borrowers

1,420 14,000 531

No. of savers

3,400 No voluntary savings No savings in Small Business Unit

Savings balances

Minimum US$10 Average US$35

- -

Financial performance

Older branches profitable, total unit almost

80 percent operational self-sufficiency

Profitable

Sources: interviews with programme managers. Note: data for all three institution refers to latest available in April/May 1999.

Loans are made to individuals within a group and all members must have individual

‘community savings accounts’. The Commercial Bank of Zimbabwe charges its community

banking clients an interest rate of 47 percent on a declining basis, while their deposits attract

19 percent.4 Arrears are standing at only 0.02 percent, and the Community Banking Unit has

3 It should be noted that the term NGO is applied loosely here. Legally, the Zambuko Trust is established as a private company, not as a charity. Due to its objectives and activities, it is appropriate to subsume it under the NGO sector. 4 The interest rates are similar to those used elsewhere in the Commercial Bank of Zimbabwe. However, non-community borrowers do not have to hold 20 percent of their loan amounts in low-interest bearing savings accounts. The nominal interest charge, however, is now slightly negative as Zimbabwe’s inflation rate edged above the 50 percent mark in early 1999.

11

not yet had any claim to the Credit Guarantee Company. DFID substantially subsidises the

unit’s operational costs and therefore meets most of the costs of credit extension.5

The management of the Community Banking Unit is convinced of their advantageous

position as part of a commercial bank. According to this assessment, microfinance NGOs are

not sufficiently concerned about the quality of their portfolio and fail to implement good loan

assessment procedures. In fact, the Community Banking Unit applies credit assessment

procedures that are surprisingly similar to those used by microfinance NGOs and it also

enjoys a donor-financed 80 percent cash guarantee cover for new clients under the

community banking scheme, covering about 40 percent of their Community Banking

portfolio.

In credit administration, the Commercial Bank of Zimbabwe appears different from NGOs.

Although group and centre meetings exist as with other Grameen-model lenders, meetings

are less frequent and less rigorously enforced. Monthly meetings have a low attendance rate

and the bank acknowledges that informal business people should not be forced to spend

valuable time in frequent and lengthy meetings.6 Loan disbursals and repayments are also

organised in a bank-like manner. Following a letter of offer for a loan, the appropriate

amount is deposited at branch level into the client’s savings account. Repayments are made

from the savings account with colour coded pay slips. The Community Banking Unit benefits

from the general infrastructure, the staff establishment and skills of the bank at large, giving

5 As part of the agreement with DFID, 75 percent of staff costs are reimbursed as well as all the costs of computer hardware, maintenance of all equipment, and the running costs of motorcycles used by credit officers. The motorcycles themselves were provided by Care International. 6 The imputed costs of time spent at meetings can add significantly to the costs of Grameen-style loans. Elsewhere I have shown that these costs can be higher than the monetary interest payable to the lender (Reinke, 1998b).

12

it a substantial advantage over non-bank organisations, such as NGOs, wishing to establish

similar systems.

The Commercial Bank of Zimbabwe is bullish about the community banking programme and

its long-term success.7 There is no doubt that the bank has become a major microfinance

provider in its country, significant both for its portfolio size and its institutional development,

although some critics in the Zimbabwean microfinance sector query its financial

performance. As a result of its entry into this previously NGO dominated field, the

Commercial Bank of Zimbabwe, sees itself in competition with NGOs rather than as a

potential collaborator. This fact is highlighted by the fact that the microfinance department at

the Commercial Bank of Zimbabwe employs several former staff members of the Zambuko

Trust, the largest microfinance NGO in Zimbabwe. According to Zambuko’s management,

competition for experienced and trained staff as well as established clients has intensified due

to the bank’s expansion in this market segment.

Zambuko Trust (Pvt) Ltd.

Zambuko Trust is the largest microfinance NGO in Zimbabwe. It is a Christian organisation

with a significant interest in the economic, as well as spiritual, well-being of its clients.

Registered in 1990, Zambuko – meaning ‘bridge’ in Shona – became operational in 1992. For

regulatory reasons, it was established as a private limited company owned by the Gesher

Trust, a Christian social welfare organisation and is licensed as a moneylender under the

7 Like many other group lenders, the Commercial Bank of Zimbabwe now explores possibilities for setting up an intermediate loan scheme for individuals who have outgrown the group based community banking scheme. To make this feasible, a way has to be found to ‘individualise’ the group guarantee system. The long-term prospect of the Community Banking Unit hinges, though, partially on the implications of the recent take-over of the Commercial Bank of Zimbabwe by ABSA, a large South African banking group.

13

moneylending act.8 The Gesher Trust started the Zambuko Trust as a response to the adverse

effects that the Economic and Social Adjustment Programme – structural adjustment in

Zimbabwe – had on poor people (Fidler and Malhotra, 1999).

Zambuko offers three loan products, akin to the three main ‘models’ in microfinance. Their

most important scheme offers group credit similar to Grameen Bank replications elsewhere.

A poverty-focused programme offers loans to ‘trust banks’, similar to the village bank model.

In addition, a small number of loans are also given to individuals. Loans are generally given

as working capital, but some clients with good track record have been awarded loans for

fixed assets. Individual borrowers get the largest initial loans of up to Z$ 8,000 (US$ 200),

while the comparable figure is Z$ 5,000 for group borrowers and Z$ 2,500 for trust bank

members. For all successive loans, all borrowers may keep doubling the previous loan

amount if their credit assessment is positive. The upper credit level is Z$ 100,000. Individuals

have to provide personal guarantees on the basis of fixed income (guarantor and borrower are

usually not the same person) while both groups and trust banks use mutual guarantees. Trust

banks are constituted of poor women, most of who are illiterate. In the design of products, the

training component and its client composition, the trust banks are the most poverty-oriented

product in Zambuko’s portfolio. It is mainly in their group and individual loan programmes

that Zambuko products compete with those of the Commercial Bank of Zimbabwe, while

mature individual clients could reach the lower levels of Barclays Small Business Unit.

8 Most microfinance institutions in Zimbabwe are registered as moneylenders under the Moneylending and Rates of Interest Act of 1996. A moneylender license does not permit deposit taking activities, but allows a regular credit business with minimal reporting requirements. Like banks, moneylenders are supervised by the Reserve Bank of Zimbabwe; the license is renewable annually (Government of Zimbabwe, 1996). A full banking license is the prerequisite for deposit taking activities, but its statuary requirements keep it out of reach for microfinance institutions for the immediate future.

14

In early 1999, Zambuko had 14,000 active borrowers, mostly group scheme participants. Its

total loan portfolio amounts to Z$ 26 million, generated through 24 branches. As part of its

programme, Zambuko also takes savings, but for regulatory reasons – ie, for lack of a

banking license – all deposits have to be invested at licensed institutions. The largest annual

write-off was three percent, while arrears are around ten percent.9 According to its

management, Zambuko is now 80 percent operational self-sufficient and May 1999 was the

first full month showing full cost recovery. Once full sustainability has been established,

Zambuko plans to acquire a banking license allowing it to develop products with a stronger

savings element and lower transaction costs.

Initially, Zambuko started with plans for cooperation with commercial banks, including

graduating its own clients to the commercial banking sector. In principle, this option has

never been dropped but stands in contrast to its strategy of transforming into a commercial

bank within the next few years. To achieve and safeguard its own profitability, Zambuko tries

to protect its client data and hold on especially to good, mature clients.

Its only tangible form of cooperation with commercial banks is an overdraft facility that

Zambuko enjoys at Barclays Bank. However, this is an expensive source of funding and

Zambuko does not rely on it for long-term finance. Barclays Bank has its own Small Business

Unit, providing small loans to established firms with the support of donor-financed credit

guarantee schemes. As this unit does not provide micro-level finance, there is only limited

competition with the Zambuko Trust (Barclays Bank, 1998). In the long run, though, more

9 It should be noted, though, that Zambuko has been criticised for writing off bad loans too cautiously. Therefore, the real default rate may well be higher than the one stated here. Also, information on arrears is insufficient; on-time repayment is said to be between 84 and 90 percent. Until 1996, Zambuko had no write-off policy. Zambuko’s management justifies their lenient write-off policy on the grounds of their clients’ repayment behaviour, who are said to be habitual late payers. A stricter write-off policy would therefore result in unreasonable write-offs.

15

direct competition is likely as the Zambuko Trust plans to provide larger, individual loans to

its maturing clientele. The relationship between Barclays Bank and the Zambuko Trust is,

therefore, one between supplier and buyer of services, not one between long-term

collaborators. Planning to acquire a banking license in the medium term, Zambuko’s

management does not pursue linkages with any other bank.10

In conclusion, NGOs and commercial banks established a relationship in Zimbabwe not by

building relationships on their respective strengths, but by moving into the same market

segment. Of course, different characteristics survive: NGOs such as Zambuko have the

widest outreach and lend to poorer people than commercial lenders but lack the institutional

sophistication found in established banks. However, up-grading NGOs and down-grading

commercial banks start vying each other as serious competitors.11 This development is not

limited to the two institutions that served as case studies here. Barclays Bank moved into

new, previously unbanked territory with its Small Business Unit, although its products are

above the micro-level. Some private sector firms, such as Nissi Finance, operate on a

moneylender license and have built up portfolios mainly in consumer lending, but are now in

the process of launching microfinance products aimed at informal enterprises. Zimbabwe’s

microfinance sector is likely to see its future development impetus coming from private

sector sources and from market-oriented agencies rather than from the traditional donors

camp (International Capital Corporation, 1999). Therefore, competition between diverse

10 The relationship between Zambuko Trust and the Commercial Bank of Zimbabwe is especially precarious. The new MD of Zambuko Trust argues: ‘Note that 95 percent of staff at CBZ [ie, community banking unit] are from Zambuko. We trained their staff. They are Zambuko Trust trading as CBZ.’ 11 The only significant ‘linkage’ project is being developed with the ILO’s Start Your Business programme. Under this scheme, the ILO will provide business training to prospective and potential clients of various lending programmes, including both banks and NGOs. Bank managers welcome this arrangement as they believe it will lead to better credit application being submitted, and that ILO graduates will be better focused and knowledgeable borrowers. As part of this scheme, ILO plans to institute a referral system. After receiving training, borrowers would be advised about an appropriate financial institution for their needs.

16

institutions is set to develop, while at the same time the distinction between NGO and

commercial bank is being blurred.

4. Government and regulation: preparing the playing field for the microfinance sector

Competitive markets are a complex net of institutions that have yet to develop in the

microfinance sectors of Southern Africa. First, markets need a variety of suppliers offering

differentiated, yet substitutable products. Although such microfinance institutions are

developing, strong, market-oriented MFIs are far and few between. Second, competition,

especially in financial markets, depends on regulation and adequate supervision. This section

illustrates the problems these requirements pose for policy-makers: regulations requires an

arms-length approach, while the institutional development needs of the microfinance sector

call for a much more interventionist strategy. It starts by outlining the new role of the Bank of

Tanzania as a promoter and regulator of microfinance in the country. It moves on to evaluate

the institution earmarked as the future flagship microfinance provider, the government-owned

National Microfinance Bank. Finally, the government-sector approaches are contrasted with

the largest microfinance institution in the country, Pride Tanzania. This section finds the lack

of supply-side diversity limits the value of regulatory efforts.

Bank of Tanzania – Microfinance Unit

As one of the first central banks in the SADC region, the Bank of Tanzania has started a

dedicated microfinance unit within its research department. This unit develops proposals and

policy guidelines for the development and regulation of a future microfinance industry and

also seeks to steer the political process and public debate on this issue.12 The microfinance

12 According to the objectives defined by government and other stakeholders, microfinance in Tanzania is to comprise savings, credit, and payment services (PriceWaterhouseCoopers, 1998, p. 5).

17

activities at the Bank of Tanzania were started with funding provided by the Japanese

government, and channelled through the World Bank, for a demand study on microfinance

services in the country. K-REP, the Kenyan microfinance organisation, was subsequently

contracted to carry out this study which evolved to comprise three separate objectives: it

sought to ascertain the level of demand for microfinance services, develop an institutional

framework for a microfinance sector and propose adequate training procedures for future

microfinance institutions. The objectives of the microfinance activities of the Bank of

Tanzania are the promotion of financial outreach into rural areas as well as a contribution to

economic growth. It is committed to promoting the use of ‘best practices’, the application of

commercial interest rates and to minimising delinquencies. The ultimate objective is to

integrate a competitive microfinance sector into commercial finance without compromising

its specific features. A separate regulatory framework will play an important role in achieving

this goal.

In pursuing its objectives, the microfinance unit works together with potential future and

present microfinance institutions, donors and policy makers (Rumambey and Thompson,

1997). It has held a seminar for members of the Tanzanian parliament in order to explain its

objectives, the nature of microfinance (especially in contrast to old-style rural finance

programmes), and to promote such ideas amongst opinion-makers in rural areas (Mugo,

1999). It has further held or attended further workshops to stimulate informed debate about

microfinance between government, donors, and potential microfinance institutions. Most

important, it remains involved in capacity building efforts jointly with donors and the

voluntary sector.

18

The microfinance unit also plans to establish a ‘focal point’ for microfinance activities in

Tanzania which is to be hosted by the Bank of Tanzania. This focal point will link donors and

government to coordinate their respective activities in microfinance. Preparation for setting

up the focal point are relatively advanced, and work on all other projects of the microfinance

unit, including regulatory issues, continues. The location of the focal point, together with the

role played by the microfinance unit, confirms that the ‘[Bank of Tanzania] is being given

overall responsibility for rural and microfinance policy’ (PriceWaterhouseCoopers, 1998, p.

8). This accumulation of responsibilities – political, promotional and supervisory – could well

lead to conflicts of interest.

In summary, the Tanzanian government and the Bank of Tanzania have embarked on a

strategy that emphasises the need for a conducive institutional and regulatory environment for

the development of the microfinance sector. While one may disagree with individual aspects

of their chosen strategy, it is certainly admirable in its ordo-liberal underpinnings and

credible market-orientation. However, the relatively advanced nature of this approach is

compromised by the absence of a microfinance sector to speak of. While there are several

institutions involved in microfinance activities in some form, only two play any significant

role so far. One of them, the ‘national champion’ from the government’s point of view, is the

National Microfinance Bank. It was formed from the residual of the old socialist-era

government bank following financial sector reform in the early to mid-1990s. Despite its

alluring name, it provides only very limited microfinance services and no credit as yet. The

second, Pride Tanzania, is by far the largest microlender in the country and dominates the

microcredit sector single-handedly. Donors and government are therefore challenged to

support other potential microfinance institutions as ‘capacity building is required to bring

them to the level’ (PriceWaterhouseCoopers, 1998, p.11) where they could effectively

19

provide financial services. Without an institutionally developed, privately controlled

microfinance sector, public policy on regulatory issues will not be able to deliver.

National Microfinance Bank

Following the break up of the former National Bank of Commerce – the monolithic

government owned bank – in 1997, and the re-structuring of its assets, the National

Microfinance Bank was created as a dedicated institution for micro-savers and credit.13 The

National Microfinance Bank started life with a finance gap – negative net worth – of about

Tsh 84 billion (about US$ 120 million). As a result, the National Microfinance Bank cannot

have a banking license and operates under a memorandum of understanding between its

board of directors, the ministry of finance, and the Bank of Tanzania. This memorandum

allows the National Microfinance Bank to take deposits, but includes several constraints to its

business. Therefore, the National Microfinance Bank is not allowed to extend new loans or

introduce new products. As its management has only limited autonomy, the National

Microfinance Bank is meant to be closely supervised by the Bank of Tanzania. In order to

allow the National Microfinance Bank to operate, the government of Tanzania has provided

an interest-bearing bond for Tsh 69 billion for bridging the finance gap. As a successor to the

main government bank, the National Microfinance Bank operates most government accounts

and all wages paid by the government sector pass through it. Mainly as a result of its

government-related business, the National Microfinance Bank has a deposit base of about

Tsh 159 billion (almost US$240 million).

13 On 1 October 1997, the National Bank of Commerce was split into three separate institutions: the National Bank of Commerce (1997) Ltd – a fully fledged commercial bank, the NBC Holding Corporation – managing liabilities and assets of the old NBC, and the National Microfinance Bank – to be developed as a fully fledged microfinance bank.

20

Savings accounts at the National Microfinance Bank pay 7.0 percent interest, while the assets

of the institution comprise treasury bills with interest rates between 9.6 and 13.0 percent. As

the national inflation rate stands at about 9.0 percent, deposit rates are negative in real terms.

Its main attraction for clients, though, is the low opening balance: the minimum opening

balance at the National Microfinance Bank starts at Tsh 3.000 (depending on branch

location), while most other banks require Tsh 50.000. Interest is being earned on balances

above Tsh 20.000. Operating the most extensive payment system in the country, and handling

a large number of small transaction and deposits through its network of 95 branches across

Tanzania, is the National Microfinance Bank’s only claim at present on being involved in

microfinance. However, it has no products specifically aimed at the microenterprise sector

and provides no credit.

The current set-up – without a license and under a restrictive memorandum of understanding

– circumscribes the freedom the National Microfinance Bank has in restructuring and

developing its business. While senior management stresses the need for additional funding,

including the expansion of its deposit base, the present arrangements prohibit advertisements.

Also, a pilot micro-lending project was planned to start during the first half of 1999, but the

government has delayed giving its approval. New savings products, some involving up-to-

date technology and smart cards, are being developed, but their introduction is questionable

as the memorandum restricts capital expenditure. Finally, as a not yet privatised former

parastatal, the National Microfinance Bank also falls under the auspices of the Parastatal

Sector Reform Commission, a body overseeing the restructuring and privatisation process in

Tanzania. This additional level of supervision imposes its own restrictions on investment,

borrowing and recruitment. Due to the bank’s history, its staff is insufficiently trained for

fulfilling their role as innovative microfinance promoter. The Tanzanian government,

21

recognising this problem, seeks to contract external agencies for managing the new

institution. Also, the institution sought to acquire relevant skills by placing a small number of

staff on internments with Pride Tanzania, a microfinance NGO. For large-scale reform like

training programmes, retrenchment and recruitment schemes, the bank has neither the

resources nor the mandate.

The restrictions imposed on the National Microfinance Bank are prudent from a central

bank’s view, although they are incompatible with the objective of spearheading developments

in microfinance. Microfinance institutions are invariably innovative and experimental, they

need to take risk and need exceptional freedom. The National Microfinance Bank needs to

protect its deposits and maintain its payment system through a period of painful reform. It is,

therefore, ill suited to play its role as a national microfinance bank as it is neither a

microfinance institution nor a licensed bank, nor is it well placed to acquire relevant

expertise. The occupation with the National Microfinance Bank has, unfortunately, distracted

attention from the need to develop an independent, diverse microfinance sector.

Pride (Tanzania)

Tanzania is a latecomer to the microfinance field. Institutional problems, including instability

in the financial system and its reform in the early 1990s, account for some delay in

developing microfinance, together with a long tradition in Tanzania of welfare rather than

market-oriented donor intervention. Today, there are a number of organisations providing

microcredit as part of their portfolio of services in business training, community development

and other areas. However, Pride Tanzania there is only one large and significant microfinance

institution so far.14

14 One other microfinance institution at least, the Gatsby Trust, is performing well but still too small to be considered significant. Previously government owned banks have withdrawn from small-scale lending and

22

Pride Tanzania is a Grameen Bank replicator and a late arrival in Pride Africa’s network of

microfinance institutions.15 It was started in 1994 and started disbursing loans in May that

year. In mid-1999 it had almost 27,000 active loan clients and a loan portfolio of just under

US$ 3 million, making it the largest microfinance institution by outreach in East Africa, and

one of the largest in Africa. Its first branch was in Arusha, where Pride still maintains its

national head quarter, but now it operates 22 branches around Tanzania. Loans are granted to

individual borrowers organised in mutual liability groups. Repayments are on a weekly basis

while interest is charged at 30 percent per annum flat. In addition, there is a forced savings

element – a contribution to a loan insurance fund – at Tsh 1,000 per week throughout the

period of membership. This amount can be withdrawn on exit; it attracts an interest ‘bonus’

of 10 percent per annum flat.16 Weekly meetings take about one hour and all financial

transactions are carried out during this time. Pride’s products and procedures are typical for a

Grameen replicator.

Table 3 - core data of Pride Tanzania

Portfolio size US$2.94 million No. of clients 26,744 Loan size Minimum US$72

Maximum US$1,500 Average US$110

Savings component No voluntary savings Obligatory contribution of US$1.5 per week, earns interest and can be withdrawn on exit from programme

Financial performance Operational self-sufficiency at 77.4 percent in June 1999, up from 58.4 percent in 1998.

Source: Pride (1999) and interview with Pride Tanzania. Note: all data refers to end of June 1999.

reduced their branch network while cooperative societies, once a pillar of small scale intermediation in Tanzania, have mostly dormant credit portfolios (Oketch et al, 1997). 15 Pride Africa is an American NGO offering support and consultancy services to African microfinance institutions operating under the Pride ‘trademark’ and using the Pride model, based on the Grameen Bank approach. 16 As Pride is not a licensed bank, it may not pay interest; therefore, the term ‘bonus’ is applied to the yield of balances held in the forced savings fund. Pride also plans to change the regulation that prohibits members from making withdrawals except when leaving the scheme. This has forced willing members to exit in times of cash flow problems and led to unnecessary membership losses.

23

However, in order to achieve its outreach, second only to the previously government-owned

and subsidised banks, Pride Tanzania has established a sophisticated management

information system, well adapted to local conditions. Its loan tracking system is still

centralised, pending computerisation and training at regional level. Therefore, records from

all client meetings have to be sent to Arusha daily, where records are up-dated and

disbursement orders are prepared. An elaborate system for sending mailbags has been

established, relying mainly on collaboration with long-distance bus companies, but also

involving fax transmissions and trains to reach its most western branches and planes for those

in Zanzibar and at Lake Victoria. This system guarantees overnight despatch of information

from most branches, two-day despatch times for all others. Importantly, repayment records

are up-dated in Arusha and disbursement information returned to branches before the next

group meeting.

Considering the enormous challenges that Tanzania poses in establishing a nation wide

network of branches, due to unreliable infrastructure and the sheer remoteness of some

towns, this system gives Pride a lead over any potential competitor. The opening of new

branches is on hold for the time being, as Pride seeks to improve its financial performance by

expanding its client base from existing branches and through new products. Particularly,

Pride explores its options for introducing an ‘intermediate’ loan product aimed at those

clients whose credit capacity has outgrown its existing credit scales. Such products would

move Pride closer to the product range available from the large number of commercial banks

in Tanzania. At the same time, Pride management seeks to enhance the efficiency of its

production cycle through computerisation of branches and subsequent decentralisation. Also,

it looks at reducing the standard time for client meetings in order to improve the productivity

24

of field staff. According to its current projections, it will be fully self-sufficient in 2001 and

will then apply for a microfinance license from the Bank of Tanzania under regulation that is

still to be developed.

In summary, Pride Tanzania has moved way ahead of its current or future competitors in the

microfinance field. While small in comparison to microfinance institutions in Bolivia or

Bangladesh, and far from profitability, Pride Tanzania is one of the most advanced

microfinance institutions in the SADC region in terms of outreach, institution-building and

future growth potential. While it needs to address various problem areas – mainly financial

performance and product development – it has the potential for dominating the Tanzanian

microfinance sector for many years. This highlights the need for donors, policy makers and

the private investors to promote credible institutions as additional players in the microfinance

sector (Nyalusa et al, 1997). These are likely to be up-graded NGOs or purpose-planned new

institutions – the problems of the commercial banking sector in Tanzania seem to limit the

possibilities of involving existing banks (Oketch, 1997).17 Despite its good start, the success

of the Bank of Tanzania in supporting a microfinance sector integrated into the national

financial system hinges on private-sector responses.

5. Conclusion: the relationship between market and microfinance

In section 2, two alternative views towards developing the microfinance sector in Southern

Africa were outlined: classical linkage development with the commercial banking sector, as

seen in the example of GAPI from Mozambique, or the full institutional integration of

17 Tanzania has a rather large number of licensed commercial banks, not all of which appear without fault. Since liberalising the financial sector in 1993, the Bank of Tanzania had to take over banks on three occasions; the last

25

microfinance organisations into mainstream finance, possibly through intermediate regulatory

arrangements, as illustrated by the rural banks in Ghana.

On first sight, the linkage model appears practical, implementable and potentially successful.

However, the example of GAPI from Mozambique stands out not only for its relative success,

but also because the conditions on which it relied are unlikely to be met in other

circumstances. GAPI lends larger amounts to its clients than all other institutions discussed

here. Therefore, its working practice is more familiar to its partner bank and their regional

extension staff and less like in a typical microfinance environment. Maybe even more

important, its partner bank had a strong developmental ethos that will probably not be viable

in a more competitive banking environment, in Mozambique or elsewhere.

The case studies from Zimbabwe illustrate the potential conflict between linkage

development and mainstreaming. Only if MFIs remain sufficiently different and distanced

from banks – especially in terms of their target group and product range – can linkages be

mutually attractive. If parallel down-gearing and up-grading exercises bring MFIs and banks

into partial competition, cooperation will make room for competitive positioning. The fact

that banks deal with small enterprises, and that MFIs plot their own mainstreaming

programme, is per se positive from a policy perspective. However, donors should be aware

that the institution building requirements involved in mainstreaming microfinance are much

more onerous than those of ad hoc linkages between banks and MFIs.

Tanzania sets an example for the SADC region with its initiative for developing a regulatory

framework for microfinance. However, the ordo-liberal approach of the central bank should

incident happened in April 1999. Some of these problems seem to be spill-overs from rather larger banking sector problems in Kenya and Uganda.

26

be accompanied by a market-oriented promotion of the supply side. Diversity and

competition in the microfinance sector are beneficial for the sectors’ development, and for its

clients, yet it will not be achieved through the establishment of a conducive environment

alone. This lesson, in particular, seems to hold some value for other SADC countries, but also

illustrates the problems for the few countries where market-oriented providers are unlikely to

appear due to the problems of remoteness and market size (Buescher, 1999).

Despite the cautionary note on linkages, the relationships MFIs maintain with commercial

banks remain important. All MFIs, except the most insular, maintain linkages in some form:

they use the services of banks for internal as well as international transfers, they keep

reserves and liquid assets in accounts or money market instruments, and they may have

access to overdraft facilities. Managing these relationships well is important, although they

are more like the relationship with a service provider than with a strategic partner. Such links

work best if the services received are fully paid for at market or cost-determined rates. MFIs

are also well advised not to share their information too freely with partner banks, especially

on their client base, and to provide appropriate incentives, through pay schedules and career

development, for retaining their most productive staff.

The process of mainstreaming microfinance in Southern Africa poses many further questions:

what is the appropriate legal status of MFIs in the formal sector? What role should

intermediate institutions, such as Ghana’s rural banks, play in the graduation of MFIs (a

question directed at regulators as well as other stakeholders)? How can MFIs best preserve

their target group orientation? What are the managerial, organisational and legal steps that

MFIs’ management and donors should follow in graduating an NGO into a formal financial

institution? These questions are being debated in Southern Africa on a case-by-case, ad hoc

27

basis. However, SEPAC and other regional groups have a role to play in formulating generic,

yet adaptable answers. Further research, more directly focused on institutional graduation,

will be necessary to support the formulation of user-oriented, practical answers for wide

dissemination.

References:

Adams, D., D. Graham and JD Von Pischke, 1984, Undermining Rural Development with

Cheap Credit, Westview Press, Boulder.

Barclays Bank, 1998, Small Business Loans – Client Information, Harare, mimeo.

Braverman, A. and J. Guasch, 1994, ‘Administrative Failures in Government Credit

Programmes’ in K. Hoff, A. Braverman and J. Stiglitz (eds) The Economics of Rural

Organization: Theory, Practice and Policy, Washington DC, The World Bank.

Buescher, L., 1999, Institutional Credit/Finance for Small and Micro Businesses in

Botswana, SEPAC, Gaborone, mimeo.

Chidzero, A-M., C. Froehlich-Lassimo, H. Hunguana and C. Cuevas, 1998, Mozambique

Microfinance Study, April, INDER, Maputo.

Churchill, C.F., 1997, ‘The Shifting Paradigm in Microfinance – the case of Get Ahead’s

Stokvel Lending Programme’, Small Enterprise Development, Vol. 8, No. 1, pp. 32-40.

Commercial Bank of Zimbabwe, 1998, description of own microfinance activities, untitled,

mimeo.

Fidler, P. and M. Malhotra, 1999, ‘Zambuko Trust’, in S.A. Breth (ed.), Microfinance in

Africa, Sasahawa Africa Association, Mexico City, pp. 172-80.

28

Fuglesang and Chandler, 1986, Participation as Process: What We Can Learn from the

Grameen Bank, Bangladesh, NORAD, Oslo.

Government of Zimbabwe, 1996, Moneylending and Rates of Interest Act, Government

Printer, Harare.

Gudger, M., 1998, Credit Guarantees – An assessment of the state of knowledge and new

avenues of research, FAO, Rome.

Hulme, D. and P. Mosley, 1997, ‘Finance for the Poor or the Poorest? Financial Innovation,

Poverty and Vulnerability’, in G. Wood and I. Sharif, Who Needs Credit? Poverty and

Finance in Bangladesh, Zed Books, London.

International Capital Corporation, 1999, Supporting a viable and sustainable microfinance

industry in Zimbabwe – Review of the current situation and suggestions for a way forward,

Open Society Initiative for Southern Africa and International Capital Corporation, Harare.

McGuire, P.B. and J.D. Conroy, 1997, ‘Bank-NGO linkages and the transaction costs of

lending to the poor through groups’, Small Enterprise Development, Vol. 8, No. 1, pp. 4-15.

Mosley, P., 1996, ‘Metamorphosis from NGO to Commercial Bank’, in D. Hulme and P.

Mosley (eds), Finance against Poverty, Vol. 2, Routledge, London, pp. 1-31.

Mugo, M., 1999, Microfinance Workshop for Parliamentarians Dodoma 22-23 January

1999, summary, Microfinance Unit, Bank of Tanzania, Dar es Salaam.

Ndyalusa, S. Mbusa, N. Ramsick, S Rusibamayila and M. Stephens, 1997, Rural and Micro

Finance Mission to Tanzania: The Role of NGOs, in: ‘Rural and Micro Finance in Tanzania’,

Joint Government of Tanzania and Donor Review, final reports, mimeo.

Oketch, H.O., J.K. Ndulu and B.M. Nyagetera, 1997, Survey and Analysis of Rural/Micro

Financial Institutions in Tanzania, K-REP and Bank of Tanzania, Dar es Salaam.

29

Otero, M. and E. Rhyne (eds), 1994, The New World of Microenterprise Finance,

Intermediate Technology Publications, London.

Owens, J. and S. Wisniwski, 1999, Microsavings: What We Can Learn from Informal

Savings Schemes, paper presented at the Third Annual Seminar on New Development

Finance, Frankfurt, 30 September.

PriceWaterhouseCoopers, 1998, ‘Microfinance Policy Consultation’, Private Sector

Consultancy Services, volume 4, November, USAID/Tanzania, Dar es Salaam.

Pride, 1999, Pride Tanzania: Facts and Figures, at http://www.prideafrica.com.

Reinke, J., 1998a, ‘How to Lend like Mad and Make a Profit: A Microfinance Paradigm

versus the Start-Up Fund in South Africa’, Journal of Development Studies, Vol. 34, No. 3,

pp. 44-61.

Reinke, J., 1998b, ‘Does Solidarity Pay? The Case of the Small Enterprise Foundation, South

Africa,’ Development and Change, Vol. 29, No. 3, pp. 553-76.

Rubambey, G. and T. Thompson, 1997, Re: Submission of K-REP Final Reports, aide

memoir, Bank of Tanzania, Dar es Salaam, mimeo.

Seibel, H.D., 1989, ‘Linking informal and formal financial institutions in Africa and Asia’, in

J. Levitsky (ed.), Microenterprises in Developing Countries, Intermediate Technology

Publications, London, pp. 97-118.

Soyibo, A., 1996, Financial Linkage and Development in Sub-Saharan Africa: The Role of

Formal Financial Institutions in Nigeria, Overseas Development Institute, London.

30